Digital Commons Network^™

Generating Audio Narration For Streaming Video Using Large Language Model, Ágoston Weisz, Andrew Goodman

Defensive Publications Series

Visually impaired individuals can find it difficult to understand their surroundings as they navigate the world. In some cases, individuals may be in situations where they are unable to look at their surroundings but need to understand their context. This disclosure leverages the capabilities of large language models to perform video understanding tasks to assist users that may not be able to view their surroundings. Per the techniques, video captured by a user device is provided to a multimodal model. A narration prompt is provided to the multimodal model, instructing the model to generate a narration of what is depicted …

Intelligent User Tailored Battery Saving Based On Usage Patterns, Prem Edhara

Defensive Publications Series

This publication describes intelligently enabling or disabling various capabilities (e.g., health tracking, GPS tracking, voice calling, messaging, etc.) of a computing device (e.g., mobile devices, wearable devices, etc.). This publication may enable a mobile device (e.g., cellular phone, smartwatch, etc.) to increase battery life by using machine learning (ML) techniques to provide recommendations on which device capabilities should be disabled. The computing device may monitor the usage of each device capability to determine a set of historical actions associated with the device capability. For instance, the computing device may monitor the foreground usage of various applications (e.g., services and/or processes …

Reducing Device Battery Consumption Using Network Condition Analysis, Nazih Almalki, Prem Edhara, Mark Wheatley

Defensive Publications Series

This publication describes techniques for optimizing device network usage by managing networking-dependent tasks on a computing device based on network connectivity characteristics of the computing device. Mobile computing devices (e.g., smartphones, smartwatches, tablets, etc.) may use task scheduling systems to determine when to process a network-dependent background process, such as synchronizing data from an on-device application with the cloud. The mobile computing device may connect to one or more networks (e.g., WIFI, LTE, 5G, BLUETOOTH, etc.) that provide connectivity to one or more other computing devices. The mobile computing device may analyze one or more network characteristics (e.g., transmit rate, …

Method And System For Downhole Electromagnetic Telemetry, Baker Hughes Company

Defensive Publications Series

The present disclosure is directed to systems, methods, and techniques for performing downhole electromagnetic telemetry.

Defined-Trust Domain Virtual Clock, Kathleen Nichols

Defensive Publications Series

Defined-trust Domains (DtD) are a type of Limited Domain [RFC8799] where the rules specifying the networking of application information are defined in a communications schema that governs the information exchanged in a particular Limited Domain. Member identities are distributed as a chain of trust, public certificates that have each been signed by the signing key associated with the next certificate in the chain and, at the root, signed by the trust anchor of the DtD. This approach is detailed in [DTD,DTLD,IOTK,TST] and example implementations with a Defined-trust Transport (DeftT) that handles communications and presents a publish/subscribe API are available at …

Apparatus, System, And Method For Wide Fields Of Illumination In Eye-Tracking Applications, Jon Bills

Defensive Publications Series

To enlarge the field-of-illumination (FOI), we propose a monolithic emitter solution, where we add the beamshaping functions directly over the light emitters (LED or laser) thorough nano/micro patterning approaches. The typical LED or laser emitter may have a flat finish surface (semiconductor material or some anti-reflection coating layers). In our proposal, we consider directly patterning beamshaping components onto the coating layer or semiconductor material through semiconductor lithography and etch process. Alternative approach may pattern the beamshaping components via imprint or other additive nano/micro manufacturing processes. One advantage to this solution is that the beamshaping function can be engineered to control …

Apparatus, System, And Method For Performing Multi-Wavelength Infield Imaging Based On Low-Index Waveguides For Eye Tracking, Jon Bills

Defensive Publications Series

To enlarge the field of view (FOV) in an artificial reality device, we propose a multi wavelength solution which essentially use two different wavelengths and each wavelength will cover a separate portion of the FOV which was tiled in the camera. However, due to the the grating diffraction, there will have ghost imaging. The ghost imaging will decrease the SNR and cause confusion for the algorithm. To address the ghost imaging problem, we propose 1)to use a dual channel camera so that information from different wavelengths can be separated. 2) one may apply time sequential imagining for each color. For …

Apparatus, System, And Method For Generating Circular Fringe Projections For Profilometry-Based Eye Tracking, Jon Bills

Defensive Publications Series

We propose a new projector design to generate circular fringe projector for eye tracking. The projector consists of a collimated laser source, a Pancharatnam-Berry phase (PBP) lens, and a linear polarizer. The PBP lens splits incident linearly-polarized collimated laser beam into a converged(diverged) right-handed circular polarization(RCP) beam and a diverged(converged) left-handed circular polarization (LCP) beam. The 2 beams interference generates linear polarization angle distribution along radial direction. The linear polarizer convert the radial linear polarization angle distribution to radial light intensity distribution, which is a circular fringe.

Inferring Implicit Timestamps From Comments, Florian Hartmann, Duc-Hieu Tran

Defensive Publications Series

Online video sharing and social media platforms enable viewers to leave comments referring to particular timepoints in a video. However, many times viewers fail to provide a link to the timepoint within the video. This disclosure describes techniques that leverage machine learning models to infer implicit or unmentioned timestamps in a comment on a video and to automatically generate links to the timestamp. Per the techniques, the video is divided into chunks. Each chunk is embedded into a feature space. The comment is optionally broken into segments and each segment embedded is into the same space. Comment segments and video …

Ranging Protocol Selection Framework, Jin Chen, Tamas Pataky, Janghyeok Choe, Arthur Tiv

Defensive Publications Series

This publication describes determining techniques for selecting between a variety of wireless ranging protocols used to determine a distance between two computing devices. An application executing at a first computing device (e.g., mobile device, smartphone, smart watch, computerized glasses, rings, vehicle head unit etc.) may request the computing device determine the distance to a second computing device. The computing devices may determine a set of available ranging protocols (e.g., Ultra-Wideband (UWB), BLUETOOTH Channel Sounding, WIFI Round Trip Time (RTT), BLUETOOTH received signal strength indicator (RSSI), WIFI, RSSI, etc.) that both computing devices may utilize. The computing devices may determine the …

Cloud Sync Based On-Demand Cellular Connection Turn On, Sumukh Ashok Shevde, Nazih Almalki

Defensive Publications Series

This publication describes techniques for enabling a wearable device (e.g., watch, glasses, ring, etc.) to intelligently toggle cellular connectivity in response to receiving an alert from a companion device (e.g., mobile phone, table, computer, etc.) via the cloud. The wearable device may determine when the wearable device and a companion device are not directly connected (e.g., via one or more short range communication protocols) but remain indirectly connected (e.g., via a cloud-based service accessible over the Internet). While the wearable device is directly connected with the companion device, the wearable device may refrain from enabling its cellular modem (e.g., in …

Content Scheduler On A Smart Television, Aravind Reddy Ayuluri, Venkata Gangadhar Kanchu

Defensive Publications Series

Systems and techniques are proposed that provide a user with the ability to automatically schedule the watching of media content provided by multiple different media content providers on a computing device of the user based on the availability of the user as indicated by a calendar application running on a computing device of the user. The computing device may include but is not limited to a smart television (TV), a tablet computer, a smart phone, a laptop computer, or a mobile computing device. The computing device may be connected to a network and interfaced to the internet.

A user may …

Audience Reaction Videos, Nagaraj Ajay Karthik Nama, Shatakshi Gupta, Kanishka Mishra, Parantap Sharma, Saumya Rangani, Veenu Mishra, Netri

Defensive Publications Series

Generally, a movie or other video content, such as a live sporting event or television show. may be from a few hours long to several hours long. During the movie or other video content, there may be various exciting scenes to which the audience visibly reacts through sound, expression, and other explicit reactions. For example, the movie may include various exciting, horror, thrilling, surprising, and nerve-wrenching scenes, which may provoke various reactions from the audience. In general, there is no way to recollect all the experiences throughout the movie and, after the movie, the audience is left with a gist …

Ai-Enabled Federated Learning System For Privacy-Preserving Health Insurance Underwriting And Fraud Detection, Vijayalaxmi Methuku

Defensive Publications Series

This disclosure presents an AI-powered federated learning system designed to enhance health insurance underwriting and fraud detection while maintaining strict privacy standards. The proposed system enables healthcare providers, insurers, and regulatory bodies to collaboratively train machine learning models without sharing sensitive patient data. The federated architecture ensures secure data processing, bias mitigation, and real-time fraud detection. Key innovations include distributed AI training, differential privacy, blockchain-based claim tracking, reinforcement learning for policy adjustments, and NLP-driven prior authorization automation. This system optimizes risk assessment, reduces fraudulent claims, and ensures transparency in health insurance operations.

Meta-Sensors: Self-Discovery, Adaptation And Reasoning, Michael Xieyang Liu, Michael Terry, Alex Olwal, Savvas Petridis, Vivian Tsai, Alexander Fiannaca, Carrie J. Cai

Defensive Publications Series

This publication describes Meta-Sensor techniques implemented by a computing system that observes the environment and leverages artificial intelligence foundation models (e.g., multimodal large language models) to perform self-discovery and suggestion of useful sensor events to users, to contextually adapt event sensors, and to self-adjust event sensors to address the limitations in home security systems, smart home controllers, and other devices and services.

Driver Optimized Media And Dialing Controls, Jacob Monsein Kandel, Peter Li, Brett Pollack, Denis Nekliudov, Eric Chiang, Marija Skeparoski

Defensive Publications Series

This publication describes techniques for creating driver-safe interfaces of media-enabled applications for vehicular computing devices (e.g., automotive head unit, infotainment system, etc.). The vehicle computing system may also monitor the state of the vehicle to determine when the vehicle transitions from a non-active state (e.g., the vehicle is parked, not in motion, etc.) to an active driving state (e.g., the vehicle is moving, in active use, in a non-parked state, etc.). When the vehicle transitions between driving states, the computing system may determine if the driver is actively using any application providing audio playback (e.g., calls, video players, audio players, …

Three-Dimensional Mesh Editing Using Masked Large Reconstruction Models

Defensive Publications Series

We present a novel approach to mesh shape editing, building on recent progress in 3D reconstruction from multi-view images. We formulate shape editing as a conditional reconstruction problem, where the model must reconstruct the input shape with the exception of a specified 3D region, in which the geometry should be generated from the conditional signal. To this end, we train a conditional Large Reconstruction Model (LRM) for masked reconstruction, using multi-view consistent masks rendered from a randomly generated 3D occlusion, and using one clean viewpoint as the conditional signal. During inference, we manually define a 3D region to edit and …

Country List Data Sovereignty Configuration, Anthony Dawson, Louise Evans

Defensive Publications Series

Data sovereignty laws can vary across different regions, and it is important for enterprises to be able to manage the cloud services that they utilize to ensure compliance with applicable data sovereignty laws across different regions. Techniques proposed herein may provide for the ability to configure cloud services with enterprise-specific data sovereignty information, while maintaining platform flexibility in order to support operational cost reductions and also improve quality and availability, which may aid service convergence.

High Aspect Ratio Reactive Mesogens For Ar/Vr Displays, Kellie Shaw

Defensive Publications Series

An optical element includes a waveguide body extending from an input end to an output end and configured to guide light by total internal reflection from the input end to the output end, an input coupling element located proximate to the input end for coupling light into the waveguide body, and an output coupling element located proximate to the output end for coupling light out of the waveguide body, where at least one of the waveguide body, the input coupling element, and the output coupling element includes a reactive mesogen.

Adjustable Brightness And Field Of View In Artificial Reality Displays Using A Varifocal Collimator And Multi-Ridge Lasers, Kellie Shaw

Defensive Publications Series

A laser display has a dynamically adjustable field of view, brightness, and resolution. An example display includes a multi-emitter laser diode configured to emit laser light, a 2D scanner block configured to receive the emitted laser light, a fast-axis collimator disposed between the laser diode and the scanner block, and a slow-axis collimator disposed between the fast-axis collimator and the scanner block, where the slow-axis collimator includes a variable waveplate and a Pancharatnam-Berry lens.

Co-Viewing Content With Advertisements On Smart Televisions, Aravind Reddy Ayuluri, Venkata Gangadhar Kanchu

Defensive Publications Series

Systems and techniques are described that provide an enhanced and more pleasurable experience for a user when interacting with a television (TV) application running on a media player. The TV application may interface with media content providers to fetch media content for streaming to and viewing on the media player by way of the internet. The media player may be a computing device (e.g., a Smart TV, a tablet computer, a smart phone, a laptop computer, a mobile computing device, a desktop computing device, etc.) connected to a network and interfaced to the internet.

A media content viewing experience of …

Unified Media Routing For Media Streaming In A Device Ecosystem, Mario Kosmiskas

Defensive Publications Series

Media routing from mobile devices to devices such as earbuds, headphones, speakers, TVs, etc. is fragmented, leading to issues such as misrouting, high development and maintenance overhead, inconsistent behaviors, missing features, etc. This disclosure describes techniques to unify media and call routing across media handling components of a mobile operating system. A graph structure is implemented to manage the sources and destinations of media. The graph structure enables precise yet flexible routing of media to devices within a user’s device ecosystem. A single source of system-wide routing information makes media routing robust, intuitive, and immersive. A flexible and scalable architecture …

Effective Protein-Protein Interaction Exploration With Ppi Retrieval

Defensive Publications Series

Effective Protein-Protein Interaction Exploration with PPI retrieval

Dynamic Ai Generated Chat Room Avatars Based On Chat Context, Andrew Mcglynn

Defensive Publications Series

Chat/messaging applications often allow users to create avatars or icons for chat rooms, spaces, or the like, which can lead to poor visual experiences when default images are used, especially when multiple chat rooms are stacked in a list. Proposed herein are techniques to facilitate an auto-generated avatar based on the chat room's title or context. Such techniques may enhance the visual experience for users by having bespoke images per chat room.

Pip Novel Interactive Method Between Ar Glass And Laptop Display, Hp Inc

Defensive Publications Series

This concept providing the possibility to interactive well with PC screen' content through special design screen and AR glass. The method is trying to provide the user experience adoptive re-arrange glass side AR content moving with PC screen content and not shielding any active window on PC screen.

This concept providing new features and solved pain points that AR glass not working friendly with external screen.

l New concept to continuously detecting screen and content position to make it possible work together with smart glass and your screen.

l Easy to add virtual control panel/notification/helper not interference with screen content …

Method And System For Managing Micropayments Leveraging Blockchain Technology And Dao, Nandit Khosa, Megha Veer Ji Marhatta

Defensive Publications Series

The present disclosure provides a method and system for processing micropayments to tackle the inefficiencies and high costs involved in it. The method involves creating Decentralized Autonomous Organizations (DAOs) for each individual or user that joins the network. Blockchain is the underlying technology for the creation of DAOs. Thereafter, the DAO issues a number of tokens for the user for performing payment transactions. Each token represents a certain value of assets or services. Whenever a user wishes to make a micropayment, the user may send the required number of tokens to a smart contract. The smart contract automatically executes the …

Mapping Table With Bloom Filter To Improve Ssd Unaligned Write Performance, Jun Tao, Wuzi Gao

Defensive Publications Series

An increase in the mapping table unit size without corresponding increases in the size of the host logical block address unit size results in a size mismatch that can cause a data write operation to be unaligned between the LBA and a solid-state drive. The read-modify-write operations to deal with unaligned write requests can take a relatively long time, resulting in a large performance drop. This disclosure describes techniques to improve the performance of unaligned data writes to SSD storage by employing a smaller Indirect Unit (IU) table on top of the original IU table in conjunction with a Bloom …

Confirming Authority To Transfer Account Information From A Mobile Device To A Laptop Computer By Capturing An Image Of A Code Presented By The Laptop Computer, Genevieve Williams, Bo Liu, Justin Lee, Elizabeth Chiu Chak Ching, Menghan Zhang, Shiba Sheikh, Jesse Johnston, Qian Yan, Michael Sean Hansen

Defensive Publications Series

A user may desire to transfer account information, such as username, contact information, and installed applications from a source computing device, such as a smartphone, to a target computing device, such as a laptop computing device. The target computing device can generate and present a machine-readable code, such as a bar code or QR code, on a display included in the target computing device. The machine-readable code can include information to be included in an authentication process when the secure connection is established. The source computing device can capture an image of the code with a camera included in the …

System And Methods For A Unified Speaker And Microphone Design For Ultrasonics

Defensive Publications Series

Methods and systems are described for an acoustic system capable of facilitating an audio signal pathway and capture ultrasonic signals via applications and/or algorithms associated with an electronic device. In an example, systems or methods may receive, via a sensor associated with a device, an audio signal. The received audio data may be converted to audio data. The audio data may be analyzed to determine a difference in acoustic characteristics in relation to a threshold. When the difference in acoustic characteristics is above the threshold, an indication may be sent to stop emittance of an ultrasonic pulse. Conversely, when the …

Social Activity Planning, Arun Manickam

Defensive Publications Series

The present disclosure generally relates to event planning. More particularly, the present disclosure relates to enabling users of a social media platform to plan activities.